Interplay between intestinal microbiome and multimodal MRI brain biosignatures for prediction of response to anti-TNF therapy in ulcerative colitis

溃疡性结肠炎(UC)是炎症性肠病(IBD)的主要病因之一，可导致终生疾病。除了腹泻和直肠出血外，腹痛和持续的疲惫可能是一种巨大的疾病负担。抗肿瘤坏死因子治疗的引入对患者的临床效益产生了重大影响。然而，它只在一小部分患者中有效，客观的炎症控制和对症状和患者相关结果的主观中枢感知之间仍然存在差异。我们小组以前的工作表明，功能磁共振成像(FMRI)能够直观地显示IBD患者对抗肿瘤坏死因子抗体治疗的反应迅速减少的疼痛感觉。在粘膜炎症显著缓解之前，这在主要的伤害性区域和涉及疼痛体验、情绪和身体感觉的边缘区域都是可见的。这些结果可能解释其他临床症状的预期改善，并使这些患者在长期内成为治疗应答者。已有研究表明，微生物群影响肠道-脑轴，也可能导致大脑功能和行为的变化。然而，到目前为止，UC患者肠道微生物组成的变化与MRI的变化没有直接的相关性。我们假设肠道微生物组依赖的疾病亚型可以通过特定的多模式磁共振(MMRI)生物特征来反映，这可以预测抗肿瘤坏死因子的治疗效果。在我们的项目中，我们的目标是更好地了解UC患者在开始抗肿瘤坏死因子治疗后大脑、肠道微生物组成和流行疾病活动之间的相关性。基于我们的概念验证研究，使用Erlangen的理想基础设施，并以机器学习方法为支持，我们将因此确定UC患者的肠道微生物组组成对抗肿瘤坏死因子治疗的反应，并将mMRI脑数据与个别临床参数相关联，以建立患者亚组差异，以预测抗肿瘤坏死因子治疗的反应。此外，我们的目标是定义肠道微生物组特异性的MRI生物特征，以描绘单个疾病亚型，并预测UC患者对抗肿瘤坏死因子疗效的反应。总体而言，根据UC患者肠道微生物群组成的特征结合个人MRI脑生物特征确定特定的UC患者亚型，并将其与抗肿瘤坏死因子治疗的治疗反应相关联，不仅可以增加我们对病理生理疾病的了解，还可以为个性化患者分层和预测治疗反应开辟新的途径。

Ulcerative colitis (UC) is one of the major entities of inflammatory bowel diseases (IBD), causing lifelong morbidity. Apart from diarrhea and rectal bleeding, abdominal pain and ongoing fatigue may represent a great illness burden. Introduction of anti-TNF therapy has made a major impact on the clinical benefit of the patient. Nevertheless, it is only effective in a subgroup of patients and there is still a discrepancy between objective control of inflammation and subjective central perception of symptoms and patient related outcomes, respectively. Previous work from our group demonstrated that functional magnetic resonance imaging (fMRI) was able to visualize a rapid decreased pain perception in the brain of IBD patients responding to anti-TNF antibody therapy. This was visible in primary nociceptive areas and limbic areas involved in pain experience, emotions, and body sensation, well before marked alleviation of mucosal inflammation can be achieved. These results might explain the prospective improvement of other clinical symptoms and renders those patients to become therapy responders in the long run. It has been shown that the microbiome affects the gut-brain axis and might also induce changes in brain function and behavior. However, changes in the gut microbial composition have not been directly correlated with MRI changes in UC patients so far. We postulate that an intestinal microbiome dependent disease subtype is reflected by a specific multimodal MRI (mMRI) biosignature and that this can be predictive for anti-TNF therapeutic efficacy. Within our project, we aim to better understand the correlation between the brain, gut microbial composition and prevalent disease activity after initiating anti-TNF therapy in UC patients. Based on our proof-of-concept studies, using the ideal infrastructure at Erlangen and supported by machine learning approaches we will therefore characterize the intestinal microbiome composition in UC patients in regard to response to anti-TNF therapy and correlate mMRI brain data with individual clinical parameters to establish patient subgroup differentiation to predict response to anti-TNF therapy. Furthermore, we aim to define gut microbiome specific, MRI biosignatures to delineate individual disease subtypes and predict response to anti-TNF efficacy in UC patients. Overall, defining specific individual subtypes of UC patients based on characterization of their intestinal microbiome composition in conjunction with individual MRI brain bio-signature and relate it to therapeutic response to anti-TNF therapy would not only increase our pathophysiological disease understanding, but could also open new ways for individualized patient stratification and prediction of treatment response.